Coping with competing views on capital setting

26 March 2013

In the second part of the InsuranceERM/Towers Watson roundtable on capital management, participants discuss the impact of ICAS+, the Prudential Regulation Authority's view on internal models and the importance of rating agencies

InsuranceERM / Towers Watson roundtable

Participants

Graham Fulcher, practice leader, P&C UK & Ireland, Towers Watson
Ian Farr, global ERM product leader, Towers Watson
Julian Ross, chief risk officer (CRO), Talbot
Heloise Rossouw, head of Solvency II, Old Mutual
Mark Chaplin, group enterprise risk manager, Aviva
Simon Gadd, group CRO, Legal & General
Justin Skinner, enterprise risk management director, QBE European Operations
Penny Shaw, CRO, Hiscox
David Weymouth, group CRO, RSA
Andrew Pryde, CRO, Beazley

Chaired by Peter Field, InsuranceERM

Part one of this roundtable can be read here

ICAS+ raises questions over capitalisation

Field: What has been your reaction to ICAS+, the invitation from the UK Financial Services Authority (FSA) to use Solvency II work to meet the individual capital adequacy standards (ICAS)?

Fulcher: My first thought is that the ICAS+ regime is an option – in this case the option, but not obligation, to use a Solvency II model to set capital ahead of the formal introduction of Solvency II.  And if something is an option with a real chance of some firms wanting to exercise it, then it has to have some intrinsic value.

Graham Fulcher, Towers WatsonIn terms of how firms have reacted, some  have realised that although they would like to exercise this option it does have some implications and means that in some cases they are going to have to use their Solvency II models in anger to produce real numbers, earlier than they had been envisaging in light of the delays. I do not think that that applies to Lloyd's; at Lloyd's, the Solvency II models are already being used in anger, both by Lloyd's for capital setting and by individual firms for management decisions.

Ross: From a Lloyd's perspective, the thing that surprised me about ICAS+ was that we turned our ICA model off several years ago. Our Solvency II model can calculate capital on an ICA basis and a Solvency II basis. I was surprised that it appears to be a big deal and that firms might still have two models, where one only does one thing and one only does the other.

Shaw: It may become popular to go back to the ultimate capital measurement approach of a UK GAAP/IFRS balance sheet. If everybody is turning on their ICA models and comparing to a Solvency II one-year focus and Solvency II balance sheet, this may openly question the appropriateness of one-year value-at-risk and the Solvency II balance sheet.

Skinner: I hope it does. I think that Solvency II is the only regime that has the one-year reserve risk. ICAS+ is the only thing that pushes it back towards a more sensible view of measuring capital, which will be good.

Ross: At the risk of saying Lloyd's again, if you are at Lloyd's, then 'club rules' apply: you are capitalising on an ultimate basis anyway.

Graham Fulcher, Towers Watson: "Given where the Bank of England is coming from, it is going to take a far more sceptical view of internal models"

Fulcher: But if you actually try to implement it practically, you realise it is too dangerous to do. For instance on a one-year basis you could write extremely long-tailed liability risk with almost no capital at all. The chance of anything emerging in the next 12 months to change your view of reserves is almost nil. But would any firm be happy to write this business with no capital? No, I do not think that anyone would.

Chaplin: You can get into quite complicated debates as to the logical consistency of adding capital numbers calculated on a run-off basis to capital numbers calculated on a one-year risk-adjusted basis. But clearly it is no worse than where we have been through the ICA regime since 2004 and I do not see that necessarily as a particular problem.

The only hesitancy from a multinational group perspective is that obviously ICAS+ is a UK regime, and I am not quite sure how that will ultimately interact with the activities of the other supervisors within the college.

Field: What has the FSA said about the deadlines for IMAP [the internal model approval process]?

Skinner: The FSA said that they want to see you draw a line under the IMAP process round about mid-year and, at that point, an internal panel will give us pointers on what is good and what is not quite so good. Then we will move to an ICAS+ regime.

Gadd: For us, they intended to give us feedback on the IMAP through the ICAS+ process, as we are using pretty much the same model.

How the PRA will consider internal models

Field: Will there be much change when the Prudential Regulation Authority (PRA) comes in, especially in relation to things like internal models?

Skinner: One of the first things that the PRA or the Bank of England started looking at for Solvency II, was introducing the pre-corridor minimum capital requirement (pMCR, see IERM, 2 October), which in essence was saying when the internal model capital goes beyond a certain point, the PRA will start investigating you. There was a certain amount of scepticism that as soon as you have an internal model regime, people will game the system and drag down capital levels in the insurance industry. My perception is that capital levels did not drop following the ICAS regime in 2004, so this measure seems inappropriate.

Penny Shaw, HiscoxChaplin: That is not what is quoted by the FSA. The FSA claims to have done the analysis and found that ICA requirements have gone down significantly.

Fulcher: I think that it is inevitable, given where the Bank of England is coming from, it is going to take a far more sceptical view of internal models, particularly where those models produce lower capital requirements than the regulatory benchmark of the standard formula.

Pryde: I think that on the Lloyd's side, the capitalisation levels have not reduced. If anything, they have probably been more appropriate for the stage of the cycle that we are in.

Chaplin: But the capital reduction is a bit of a red herring in some ways, as I think that there has been substantial de-risking activity. So reduction of capital is a natural consequence and even evidence that the regime has appropriately incentivised more active risk management.

David Weymouth, RSA: "Regulators in some European countries that I talk to will question why we are using an internal model when the local norm will be to use standard formula"

It is clear that the PRA is thinking more broadly and learning from the bank regulation experience. I do not think that there is a rejection of internal models as providing useful information; it is just recognition that you need a broader spectrum of supervisory and risk management tools.

Ian Farr, Towers WatsonFarr: Perhaps the UK industry and the FSA underestimated the amount of work required to establish an internal model. Lots of companies were encouraged to go down that route and then found that actually it is extremely onerous for the regulator to get comfortable with the internal model and for the company to provide backup for it.

With the 'use' test, it's not genuinely going to be satisfied within a year; it will be potentially three or four years before the model is truly bedded in. Perhaps it is worth stepping back and saying, 'We have developed internal models, but now we need to think more carefully about their role, both in a business and a regulatory context.'

Weymouth: Regulators in some European countries that I talk to will question why we are using an internal model when the local norm will be to use the standard formula. They are concerned that it seems unnecessarily complicated and are slightly sceptical.

Shaw: I think the PRA should consider doing a revised standard formula as quite high on their agenda. They have ECR [enhanced capital requirement], they have QIS [quantitative impact study] figures, but where is the mid-point? It would be a bit more sensible to be more risk-based where you know the correlation structures that you have imposed based on industry analysis and based on the data, and tell companies to stay between risk categories.

The importance of rating agency capital

Field: Will rating agency capital become more important?

Skinner: For non-life commercial business, the rating agency capital has always been the important piece. Solvency II is a lost opportunity; I was hoping that rating agencies would look at Solvency II and say, 'You delivered the Solvency II model; we are now going to use the Solvency II internal models as part of our methodology.' The option is there for them to do it, but I do not think they are pushing it particularly hard. It is a bit of a shame, and Solvency II means that rating agencies may take internal models less seriously in the short term.

Julian Ross, Talbot: "I think that many people, especially in Bermuda, would see the rating agencies as de facto regulators"

Fulcher: Yes. I think that one of the downsides of Solvency II could potentially be the slowing of the convergence of rating agency capital to economic capital. The rating agencies were starting to engage with economic capital and had plans to start reviewing and using internal models more seriously in their capital assessment.

Shaw: Boards are expected to know minute detail on internal models but what about boards' understanding of other capital models – i.e. rating agency, etc? I have done an exercise recently and compared the SCR [solvency capital requirement] with various standard formulas out there – Standard & Poor's and AM Best's, and others. They are really very different, and I think communicating the key differences and discussing why they are different and who the winners and losers are would be useful, as these are used across the industry to measure credit risk exposure to other companies.

Andrew Pryde, BeazleyPryde: One of the things that I have put into the ORSA [own risk and solvency assessment] is the different capital metrics that we are measured on, and each one is slightly different. The risk is the same, but you have the standard formula, the internal model and the ratings agencies...

Shaw: They can certainly be north of 20% to 30% different.

Pryde: They can be very different. If you are a board member trying to work that out, you can end up managing your business to the worst common denominator, as it were, which is generally now the rating agencies.

Ross: I think that many people, especially in Bermuda, would see the rating agencies as de facto regulators. I know they are not regulators, but I think in many instances rating agency capital is effectively the driver of capital. For example, although we hold more than that, we would not hold less than the amount that is required to maintain our target rating.

Field: Why do rating agencies still have that power when they have come under so much criticism?

Ross: People who buy insurance generally have no other credible means of assessing the credit worthiness of the people that write it.

Weymouth: If you want to write commercial business with mid- and large-sized companies, it is very straightforward and they just will not let the business be written with you if you do not have the right rating.

Farr: The regulators in the US appear to set their rules slightly differently; rather than trying to set a relatively high target capital requirement, they set an absolute minimum requirement, with staged intervention at various points above that. Companies are aiming to hold the order of three to five times that regulatory minimum, with the rating agencies effectively setting the capital standard to maintain the ability to write business.

Andrew Pryde, Beazley: "All our models use credit ratings, so therefore you need to think through how your capital model will respond when ratings change"

Fulcher: We can all complain about the rating agency models when they rate our own businesses, but I suspect that pretty well every company around this table on the non-life side uses the rating agencies when they decide which reinsurers to place business with.

Ross: Think about reinsurance credit committees and the repercussions of their decisions; if they approve buying reinsurance from an AA-rated reinsurer and it goes bust, compare that with approving buying it from an unrated reinsurer that then goes bust.

Pryde: It was interesting when we were putting together a eurozone dashboard. You realise that all our models use credit ratings, so therefore you need to think through how your capital model will respond when ratings change.

If you have a raft of reinsurers who are downgraded, all of a sudden your capital model could swing quite significantly. I think as we build these internal models with further data points and interdependencies, we have to think those through, because they can sometimes drive business decisions unexpectedly.

Publication of economic capital

Field: Will more firms publish their economic capital?

Fulcher: Yes absolutely. As Mark [Chaplin] has outlined, Aviva has been very explicit with its assessment of economic capital, its target level of capital, the actions it has taken and the actions it intends to take as a result of those assessments. From my perspective, that is being well received by the market, because I think they can clearly see what the drivers of capital are and how this matches the decisions being taken.

Chaplin: I think it will be hard to maintain that you run your business on economic capital and that it is driving decision-making and not communicate that to the market. Certainly, we have benefited from having that conversation with our investors, and we would imagine that other firms will see that benefit and will start publishing.

Mark Chaplin, AvivaFarr: The insurance industry – particularly the life insurance industry – is struggling with its communication to investors; the more that can be communicated around risk exposures and capital requirements, the better the industry is likely to be perceived.

Pryde: We have certainly always published; I just do not see the advantage of not. I think managing capital is one of the key parts of running the business, so it seems very odd to me that there is a silence on whether you are putting capital to use within the business, or whether it is worth returning the capital to shareholders at this point in the cycle. If you are silent on that, I cannot see how investors would be able to invest in the company.

Gadd: I think it will present a problem for analysts though, because if everybody is publishing their economic capital, how do they compare them? Embedded value was often undermined because it was assumptions-based, and it was a lot simpler than insurers' economic capital models.

I think in the context of explaining how you run your business, it is a very valuable piece of disclosure. Certainly the risk exposure is probably even more valuable from the analysts' perspective, so they can effectively run their own model to take your risk exposure and understand the risk from their perspective.

A chance to revisit strategic decisions?

Field: With Solvency II being deferred, will that encourage companies to make strategic changes that they might have been putting off?

Chaplin: It's had no impact on us from a strategic perspective. We have a degree of urgency that we have injected into our transformation programme - all of this was happening with or without Solvency II.

Ross: I imagine if we were a larger, more complicated group we might have been thinking about restructuring to optimise Solvency II, but we're quite straightforward and so it has had no real impact on our strategic decisions over the last few years.

Penny Shaw, Hiscox: "I certainly saw a number of companies using Solvency II as an opportunity to financially re-engineer how actuarial, finance and modelling teams interacted and produced information"

Fulcher: Yes, some firms have had certain regulatory-type plans, some of which they were doing purely from a Solvency II point of view, some of which they wanted to do from a business sense but thought, 'There is no way we are going to get this done before Solvency II comes in,' and therefore had deliberately shelved them. Some of those firms are now taking their business-focused plans off those shelves and dusting them down; while at the same time considering whether to shelf the Solvency II-driven plans.

David Weymouth, RSAWeymouth: I think Solvency II has probably driven a slightly more rapid review of corporate structure. If you have a relatively complex corporate structure with many legal entities, it is not only Solvency II, but a number of things would drive you to simplify that as much as you can. Again, Solvency II was more of a catalyst than a fundamental driver of it.

Pryde: Also, it's less risky like that. If you look at insurers and banks where there have been issues, one of the most common denominators is complexity in corporate structure.

Raising capital in an uncertain environment

Field: Does the delay affect the ability to raise capital on a more regular basis?

Gadd: There might be some changes to appetite of external investors that come into the insurance market, because clearly at one stage there was such a wide range of potential outcomes in terms of the regulatory capital requirements. That was presumably putting some investors off. Even though there is now a delay, the potential range of outcomes is now much narrower, which would give confidence to third parties to come into the market and put their capital at risk.

Chaplin: There is appetite in the market for insurance company debt. These are long-term capital instruments, so within any time horizon that you are likely to issue for, you are going to have Solvency II or some replacement regime that still leaves open the question as to what will count as qualifying capital. More generally, there is a push for better-quality capital. Aviva has talked about medium-term commitments to reduce our leverage and improve the quality of our capital. I think that is a common thread; rating agencies like to see it; regulators would like to see it.

Money well spent?

Fulcher: Looking back on Solvency II, what do people think are the things that they gained the most from the money they have spent?

Gadd: We have invested in data governance, the quality of the modelling and the governance that goes around the models. All of that is going to be of value for the long term. We have probably spent more money on it than we would have otherwise done, because we had to do it under time pressure and pay expensive contractors, but the end product has definitely moved us forward.

The fact that we now focus on using the model to help us actually run the business rather than, as you say, producing lots and lots of reports for regulators is a very positive thing, I believe.

Ian Farr, Towers Watson: "Solvency II poses a lot of very valid challenges to companies' use of modelling in risk management."

Weymouth: I think it has probably helped in raising awareness of managing a balance sheet as opposed to purely the P&L [profit and loss], out into the business as a whole.  Over time, I think that is positive, because I think people will be more rounded business managers across the piece.

Pryde: I think we have benefited more from the pillar 2 side. There has obviously been some benefit on pillar 1, but given that we have been using an ICA model for a number of years, there is less marginal benefit there.

The ORSA is a CRO's valuable tool to keep the board educated on risks and capital management. Would we have had that without Solvency II? Possibly. Whether that will help stop companies going bust or not is to be decided, but I think the quality of information for the board has never been better.

Shaw: I certainly saw a number of companies using Solvency II as an opportunity to financially re-engineer how actuarial, finance, modelling and other "finance" focused teams interacted and overall drove production of financial management information used by the business. Solvency II helped make it more simple, transparent and better controlled. There were also improvements in the broader understanding of all the risks the business faces.

Farr: Solvency II poses a lot of very valid challenges to companies' use of modelling in risk management. There is a danger that models are calibrated against a small data set and that calibration is just treated as a given, whereas the reality is that judgement plays a huge part in calibration of models. That reality needs to be recognised and we need to make sure that we get the right judgements into the calibration process.

Pryde: From a strategic point of view, we have looked at two transactions, and in both cases, giving the board the usual corporate finance PowerPoint presentation as to why the transaction should go ahead, but also supported with a transactional ORSA. I do not think the board would have had both those dimensions quite so explicitly in the past, so actually I think that it is driving better strategic decisions at the point of them being made, rather than going ahead with the transaction and finding out some of the issues after.

Rossouw: In my view, having only recently joined the organisation, we have benefited most from the pillar 2 work. Our risk and capital management supports sound business decisions and allows us to attribute capital according to an accurate assessment of the economic impact of our risks.